Figure 1
From: The application of network label propagation to rank biomarkers in genome-wide Alzheimer’s data
A small bipartite graph for a hypothetical dataset with five samples and two SNPs. The five samples are represented by the nodes at the left (V), and are labeled with case or control status (+1, -1, respectively). Each SNP is represented by three nodes at the right (U) with one node for each SNP state (AA = major homozygote, Aa = heterozygote, aa = minor homozygote). Edges represent actual observations in the dataset and connect samples to the SNP states that they exhibit. Labels are allowed to propagate along edges and result in a final labeling for each node in the range (-1, +1), indicating association with case or control status.